Buying in silos, and the duplicate tools it hides
Buying in silos is not a procurement failure. It is what happens when five departments hit the same problem in five inboxes and each buys its own tool. The organization ends up paying for the same thing three times, under three contracts, and no one owns the total. A shared place to see the stack and the standards ends the cycle.
By the PartnerAZ team · Published June 6, 2026 · Updated June 9, 2026
Who this is for.
The person accountable for a municipality's software stack, who keeps finding two departments running near-identical tools that nobody compared before buying.
The problem, in plain words.
Departments work in silos. One team evaluates a vendor on its own, unaware that two other departments already looked at and rejected the same vendor for overlapping reasons. The same tool gets bought three times under three contracts, and the renewal calendar turns into a part-time job nobody signed up for.
Five inboxes, one vendor.
In the public sector, silos are how the budget is built. Each department holds its own line items, its own sign-off, and its own threshold below which no tender is required, so a purchase can clear without ever crossing a central desk. Vendors understand this structure better than most buyers do. Salesforce found in 2024 that 81 percent of sales teams now use AI, which in practice lets one vendor pitch five departments at once, each message tuned to that department's words. The silo used to slow duplication down. Now it speeds it up. Nobody wants every purchase centralized, and shared standards set upstream do not require it: each department keeps its autonomy and still sees what the others see. AI can work for the buyer too, as a deterministic score per criterion where changing a weight updates the ranking.
How pre-discovery changes it.
A shared pre-discovery layer gives every team the same view of what is already in the stack, and the same standards to weigh new tools against. Set the standards once. Let each department discover inside them. Cross-department evaluations are visible across the organization, so no team redoes the research another already did.
What it costs to leave it alone.
Duplicate tools quietly become the largest line item in the technology budget that no one owns. The total is not the licenses. It is the renewals, the integrations, the security reviews, and the staff time spent working around a stack the organization never chose on purpose.
How to decide: three cheap checks, starting with the renewal calendar.
The renewal calendar is the cheapest place to start. Ask finance for every software line item across departments, even a rough export, and circulate it; the first duplicate tends to surface fast. Before any team evaluates a new category, post one question in a shared channel: has another department looked at this in the past year. The answer costs nothing and often retires an entire evaluation. And write the organization's standards down once, as scoring criteria, before the next purchase rather than during it; criteria written mid-evaluation tend to describe the favorite vendor. If you want those standards doing more of the work, PartnerAZ lets you set them privately and have vendors apply against them blind, scored per criterion, visible to every team. One afternoon covers all three. Vendors pay to be seen. They can never pay to rank. Your list is ordered by fit, which is the whole point of it.
What the evidence shows.
The duplication is measurable, even when no single team can see it. Nobody agrees on how big the stack is, which is the point: BetterCloud's 2024 count puts the average company at 106 SaaS applications, while Zylo's 2025 count puts it at 275, two reputable counts of the same thing that land 169 apps apart, a measure of how poorly anyone sees the full stack. The buying is already collective: Forrester's 2024 report found that 89 percent of purchases span two or more departments, and Gartner found in 2025 that 74 percent of B2B buying teams show unhealthy conflict during the decision. If the buying already crosses departments, the discovery should too. Shared standards are how it does, without taking autonomy from any team.
BetterCloud (2024) and Zylo (2025): the average company runs between 106 and 275 SaaS applications, two reputable counts that land 169 apps apart. Forrester (2024) adds that 89 percent of purchases span two or more departments, so the buying is already collective.
FAQ
How do we stop buying the same tool in five departments?
Publish your live stack and your standards in one place, and route every new tool request through it. When operations asks for a survey tool that finance already piloted last quarter, the earlier discovery is already on file, and a single owner, named upstream, decides whether to add a fifth tool or extend the one you have.
What does buying in silos actually cost us?
In a mid-size organization, duplicate tools quietly become the biggest line item no one owns. Every redundant tool adds two renewals, two security reviews, and integrations that conflict. The licenses are the smallest part of the bill.
Is PartnerAZ a procurement or asset-management tool?
No. We sit before procurement and asset management; your existing procurement and renewal processes do not change. We catch duplicate tools before anyone signs a second contract.